What is restaurant_ev_charging?

restaurant_ev_charging is a small web service built for electric vehicle road trip planning. It answers a specific question: given a location, which fast chargers are nearby, and which restaurants can you walk to from each of those chargers while your car charges? The service is built with FastAPI, a Python web framework, and it combines two outside data sources. It uses OpenChargeMap to find DC fast-charging stations that support common connector types like CCS and Tesla's NACS, and it uses Geoapify to find restaurants within a walking distance you choose from each charger. On top of the plain API, the project includes an optional interactive assistant that lets you ask for a stop in plain language, such as finding somewhere to eat near chargers in a particular town within a certain walk. This assistant is powered by Google's Gemini model using function calling, meaning the AI model itself decides when to call the charger and restaurant search functions behind the scenes. There is also an optional layer that enriches restaurant results with ratings, price level, and whether a place is currently open, pulled from either Yelp Fusion, which is a paid commercial service, or Google Places, which offers a free monthly usage credit. Running it requires Python 3.10 or newer and free API keys from OpenChargeMap and Geoapify at minimum, the Gemini key is only needed if you want the conversational assistant, and the review enrichment keys are entirely optional. Setup is done by copying an example environment file, filling in the keys, then running a Makefile command to install dependencies into a virtual environment and another to start the server. The API is rate-limited to 20 requests per minute per IP address and publishes an OpenAPI schema, which the README notes is suitable for wiring into a ChatGPT-style tool integration. The project also documents how to deploy it to Railway with a few environment variables set in that platform's dashboard.